New statistical methods for multiparametric problems arising in cardiovascular research are to be developed: 1. We shall develop a model leading to estimates of the effects on disease risk of intermediate, as well as baseline, measurements of risk factors in prospective studies. Time to an event and baseline and intermediate measurements will be assumed to have a multivariate distribution of known form, but with parameters to be estimated from the observations, leading to an estimate of the hazard function, conditional and intermediate measurements. The absence of measurements following an event is to be handled by assuming the existence of an ergodic hazard function. 2. Pull-back estimates and significance tests, appropriate to clinical trials, epidemiological investigations, and computer-aided diagnosis are to be further developed for (a) the one-way classification (b) multi-way classifications and newly developed for (c) multiple end points and (d) regression estimates, including the multiple logistic, Weibull survivor functions in life tables and density estimation.